Table 1 - uploaded by Mick Silver
Content may be subject to copyright.
Source publication
Abstract: Hedonic regressions are used for residential property price index (RPPI) measurement to control for changes in the quality-mix of properties transacted. This paper consolidates the confusing array of existing approaches and methods of implementation. It further develops an innovative form of weighting at the (elementary) level of the indi...
Citations
... There is a more introductory paper/text to be written to walk the reader through the estimation of, and diagnostics for, estimated hedonic regression equations, one that may include data examples. Silver (2018) takes the reader from the estimated hedonic regression to a PPI. Silver (2018) does not adopt the approach to PPI measurement that decomposes a PPI into its land and structure components. ...
... Silver (2018) takes the reader from the estimated hedonic regression to a PPI. Silver (2018) does not adopt the approach to PPI measurement that decomposes a PPI into its land and structure components. This is approach is well developed in Eurostat et al. (2013, chap. ...
... the three main hedonic approaches to PPI measurement: the hedonic time dummy approach, imputation, and characteristics/repricing approach. There are many alternative forms for each approach depending on (1) the functional form of the hedonic regression and aggregation; (2) the choice of reference, current or some average of the two, period(s) to estimate hedonic coefficients or hold characteristics/weights constant; (3) whether dual or single imputation is used for prices and/or weights; (4) whether a direct or indirect formulation is used; (5) the periodicity of the estimation, say monthly/quarterly/annually -chained, rolling window or fixed baskets of characteristics; and more -see De Haan and Diewert (2013) and Silver (2015Silver ( , 2018. A tweaking of the methodology is outlined in Subsections 3.1 and 3.2 to help deal with instances of sparse data. ...
Hedonic regressions are widely used and recommended for property price index (PPI) measurement. Hedonic PPIs control for changes in the quality-mix of properties transacted that can confound measures of change in average property prices. The widespread adoption of the hedonic approach is primarily due to the increasing availability, in this digital age, of electronic data on advertised and transaction prices of properties and their price-determining characteristics. Yet hedonic PPIs are only as good as the underlying estimated hedonic regressions. Regression-based measures are unusual in official economic statistics. There is little technical support in the international Handbooks and Guides for diagnostic measures and graphical plots for estimated regression equations as applied to PPIs. These diagnostics are essential to the transparency and credibility of hedonic PPI measurement. This article seeks to remedy this.
... Leventis (2008) has found differences in the autoregressive formulation of the RS model, used to weight such paired comparisons, can account for significant differences in the index results. Hedonic regression methods are now more prevalent than the RS method as a result of the increasing availability of detailed data sets of house prices and characteristics, mainly arising from the development of on-line residential property sales databases, and the development of a more sophisticated hedonic RPPI methodology (Hill and Melser (2008); Hill (2013); De Haan and Diewert (2013); Shimizu (2017), andSilver (2018) led to the development of international standards of measurement for RPPIs (Eurostat, 2013) and the widespread development of RPPIs both in terms of the number of countries and their quality of measurement (Hill et al., 2018). ...
... It is "superlative" in the sense that the index of price changes of transactions undertaken in period 0 makes symmetric use of reference and current period price information (Diewert, 1976 andBalk, 2008). Silver (2018) shows how a weighted version of equation (4) can also be formulated as can a superlative index that makes use of both base period 0 and current period t transactions as the two weighted versions of equations (3) and (4) are put together as a weighted average to form a Törnqvist hedonic imputation index for the full sample. ...
... The regular estimation of a hedonic regression every say quarter, using sparse data, leaves the index results open to bias from undue influence and other vagaries of econometric estimation. Compilers further benefit from using an extended time period for the reference period, as advocated by de Haan and Diewert (2013) and Silver (2018). Such indexes can take a quasi-superlative form as explained in Silver (2018)-Section III. ...
Commercial property price indices (CPPIs) should be based on market transaction prices. Yet in monitoring average price changes over time price data can be sparse and the properties transacted each period of a different quality-mix. Due to the heterogeneity of commercial property, CPPI measurement requires a quality-mix adjustment so that the prices of like properties are compared over time with like. An appealing way around this sparse data and quality-mix adjustment problem is to use price data on broadly the same properties over time and avoid transaction price data. Tax or investment appraisal data or market valuations of real estate investment trusts (REITs) are two commonly used alternatives. While convenient, both such series can seriously mislead macroprudential and macroeconomic-policy makers. In this overview paper we point to the deficiencies of these data sources, outline and argue for the use of hedonic methods of quality-mix adjustment that are designed to work with sparse transaction price data in thin heterogeneous commercial property markets.
... Hedonic approaches (for an overview see Silver, 2016 ) build on the assumption that the price or rent of a dwelling is determined by shadow prices associated with the dwelling's characteristics. For instance, there is a shadow price for a bathroom, an extra bedroom, and so on. ...
This paper compares two model‐based multilateral price indexes: the time‐product dummy (TPD) index and the time dummy hedonic (TDH) index, both estimated by expenditure‐share weighted least squares regression. The TPD model can be viewed as the saturated version of the underlying TDH model, and we argue that the regression residuals are “distorted toward zero” due to overfitting. We decompose the ratio of the two indexes in terms of average regression residuals of the new and disappearing items. The decomposition aims to explain the conditions under which the TPD index suffers from quality‐change bias or, more generally, lack‐of‐matching bias. An example using scanner data on packaged men's T‐shirts illustrates our framework.
These lecture notes cover old and new investment methods, regulatory and legal developments and the role of technology as a game changer in asset management.
The discussion gives the same weight to the theoretical and practical aspects of asset management.
The focus is on portfolio constructions, asset pricing foundations, factor investing and backtesting on the theoretical side. The applied chapters contain an asset management industry overview and pension funds, growth of wealth, digital asset management, big data and hedge funds.
Data and Python files and Exercises can be found in the Asset Management Project Folder
